Sablebourne · About this demo

What you're looking at

A public, deterministic demonstration of Sablebourne's AI brand-visibility intelligence platform, loaded with Nike as the active brand profile.

The short version

This demo uses fixture Claude responses: 3 runs per query across 60 days, showing improving, declining, and stable trajectories, processed through the same analysis, scoring, and recommendation engines Sablebourne uses in production.

Every number on the dashboard comes from deterministic TypeScript rules. The methodology page walks through every step end-to-end.

Why Nike?

We picked Nike because it exercises the platform across every signal we care about:

  • Strong owned-media footprint (5 owned domains), testing owned-source detection.
  • Crowded competitive category (15 competitors in the dictionary), testing mention-rank, competitor crowding, and comparison framings.
  • Both positive and negative narrative lenses (20 framing tags), including performance-led, culturally influential, and skepticism-bait framings.
  • Distinct structured vs. consumer-voice query styles, which LLMs handle very differently.

This demo is not affiliated with Nike. All data shown is illustrative and derived from publicly observable LLM output patterns; nothing here represents Nike's internal view or strategy.

What's real vs. what's simulated

Real

  • The analysis engine (regex-based presence, rank, directness, framing).
  • The scoring engine (presence / rank / directness / framing / sources).
  • The recommendation engine and taxonomy.
  • The database schema, query universe, and domain classification rules.
  • The brand-profile registry pattern: swap brand, swap everything.

Simulated / fixture

  • The Claude responses: hand-crafted 3-variant fixtures per category.
  • Run timestamps: spread across 60 days for a realistic trend view.
  • Per-query trajectory: deterministically assigned from a hash of the query id.
  • Live execution is disabled in this public deployment.

Configuration flags

PUBLIC_DEMO_MODEtrue
LIVE_EXECUTION_ENABLEDfalse
Active brandnike

What Sablebourne does in production

  • Live, scheduled Claude runs (and other frontier LLMs) on your bespoke query universe.
  • Domain classification calibrated to your category and owned media estate.
  • Custom framing dictionaries tuned with your comms, brand, and CX teams.
  • Strategic read-outs linking signal to action: what to change in the next sprint of owned content, PR, retail copy, and campaign work.
  • Integrations into your existing analytics and BI stack.

Want to see your own brand?

Sablebourne onboards new brands in under two weeks. Get in touch to discuss your query universe, the lenses that matter in your category, and what a first read-out would look like.